Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Mhbounds v.s. rbounds in stata


From   Wen Ci <ivwen.ci@gmail.com>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: Mhbounds v.s. rbounds in stata
Date   Tue, 26 Feb 2013 12:41:36 -0400

Thank you for your answers.

However, my question is why does the mhbounds which is supposed to fit better in my model with binary outcome variable produce the insignificant result when gamma is 1?

I suspect this might be because of the matching without replacement  used in the mhbound. Test.

Thank you for your help.

Wen


On 2013-02-26, at 12:26 PM, "Ariel Linden, DrPH" <ariel.linden@gmail.com> wrote:

> First off, I concur with Billy's general comments below pertaining to
> following accepted procedures for the Statalist.
> 
> As per you question regarding -rbounds- and -mhbounds- , both user-written
> programs (findit -rbounds- and -mhbounds-): If you read the help files you'd
> see that rbounds is for continuous outcomes and mhbounds is for categorical
> outcomes. Given this, you could expect to get different results.
> 
> Ariel
> 
> Date: Mon, 25 Feb 2013 12:23:20 -0400
> From: Wen Ci <ivwen.ci@gmail.com>
> Subject: st: Mhbounds v.s. rbounds in stata
> 
> Dear readers,
> I employed the mhbounds after psmatch2. My outcome variable and treatment
> variable are all binary variables and I think the mhbounds are a better
> choice compared to rbounds.
> However the mhbounds produce all insignificant results even when gamm=1,
> which is inconsistent with my matching results.
> I also tried the rbounds, which produce normal results for my matching.
> Is it because the mhbounds are designed for the matching without
> replacement? My matching is completed with replacement due to the sample
> size limitation.
> Thank you for your answer.
> Best,
> 
> Wen
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
> 
> ------------------------------
> 
> Date: Mon, 25 Feb 2013 08:33:32 -0800
> From: William Buchanan <william@williambuchanan.net>
> Subject: Re: st: Mhbounds v.s. rbounds in stata
> 
> Hi Wen,
> 
> It is unlikely that resubmitting the same question will yield any
> substantively different results from what you've already experienced. The
> first answer most would likely provide, is to read the Statalist FAQ as
> users are required to do prior to posting to the Statalist.  -psmatch2- is a
> user written command which can be installed from ssc.  Posters are asked to
> identify where user-written packages come from for several reasons (not
> limited to, but including determining whether the version said person is
> using is the most updated version of the package).  Furthermore, the FAQ is
> explicit about showing your exact syntax and Stata's exact output.  There's
> good reasoning for this and it helps provide other users with information
> that could be important to answering your question and giving you the help
> that you need.
> 
> HTH,
> Billy
> 
> 
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index